A task-specific validation of homogeneous non-linear optimisation approaches

Area of Science:

• Biomechanics
• Musculoskeletal modeling
• Human movement analysis

Background:

• Musculoskeletal models often exhibit redundancy, posing a challenge for determining unique muscle force distributions.
• Static, non-linear optimization methods are commonly employed to solve this distribution problem.

Purpose of the Study:

• To present and validate a method for assessing non-linear optimization approaches in musculoskeletal modeling.
• To evaluate the accuracy of homogeneous cost functions in predicting muscle forces under varying loading conditions.

Main Methods:

• Developed a validation method for non-linear optimization techniques using homogeneous cost functions.
• Predicted theoretical muscle forces for scaled loading conditions based on a single experimental condition.
• Compared predicted muscle forces with experimentally measured forces.

Main Results:

• Muscle force predictions were found to be scaled versions of each other only when joint loading conditions were also scaled.
• The study demonstrated that experimental muscle force sharing among synergistic muscles does not simply scale with joint loading.

Conclusions:

• Convex homogeneous non-linear optimization approaches are insufficient for accurately predicting individual muscle forces in biomechanics.
• Current methods fail to capture the complex force-sharing dynamics observed experimentally.